The trajectory of the HIV epidemic in coming decades will be determined by the degree to which we can identify infected persons and engage them in care - a point implicit in the ambitious UNAIDS 90-90-90 target, which sets 90% goals for HIV diagnosis, linkage of infected persons to sustained antiretroviral therapy (ART), and viral suppression in those treated. Meeting this target requires successful engagement of difficult to reach populations, such as people who inject drugs (PWID) and men who have sex with men (MSM), who bear a disproportionate share of the epidemic particularly in low to middle income countries. Our team is nearing completion of a multi-site cluster-randomized trial in India to assess the effectiveness of integrated care centers (ICCs) for PWID and MSM compared to usual care (fragmented service delivery). The ICCs, which separately target PWID and MSM, provide vertically integrated, evidence-based HIV prevention and treatment services in a single stigma-free setting. ICC process measures from the first year show robust uptake of HIV counseling and testing, the primary outcome for that trial, but slower than anticipated ART uptake. Ethnography with stakeholders identified travel time and loss of wages as barriers to attending ICCs regularly, as required for medical management of HIV. Demand-side interventions in public health (such as treatment incentives) can be particularly effective when paired with optimized treatment accessibility (i.e., supply). Consequently, we propose to examine whether provision of HIV care and treatment incentives to ICC clients - redeemable for food and household goods - will improve overall utilization of the clinics and downstream HIV care continuum outcomes. We propose a hybrid effectiveness-implementation design. This will include a 16- site, pair-matched cluster randomized trial to compare the effectiveness of adding of HIV care incentives to ICCs (ICC+) versus standard ICCs on HIV care continuum outcomes, including ART initiation, adherence and viral suppression (aim 1). Effectiveness will be compared at the ICC level (from a cohort of HIV-infected ART- eligible clients followed in each ICC and process measures deriving from all ICC clients) and at the community- level through a cross-sectional sample accrued via respondent-driven sampling (RDS) 2 years after initiation of the intervention. Because PWID and MSM will be sampled independently from the ICCs in the RDS, it provides an opportunity to characterize outcomes like community viral load and HIV incidence, reflecting impact within the broader PWID/MSM communities. As an exploratory sub-aim, we will use a rigorous scientific design to assess the effects of withdrawing (vs. continuing) incentives beyond the initial intervention phase. Additionally, we will characterize barriers and facilitators to implementation of the ICC and ICC+ interventions (aim 2), and determine the cost-effectiveness of the ICC+ intervention (aim 3). This proposal is strengthened by infrastructure from our prior trial, detailed data on the target populations in the 16 sites for the proposed tria, and an expanded multidisciplinary team with new expertise in implementation science and cost-effectiveness.